Source: UNIV OF CALIFORNIA submitted to NRP
RAPID RESPONSE TO EXTREME WEATHER TO PROMOTE HUMAN HEALTH AND AGROECOSYSTEM RESILIENCE IN THE SAN JOAQUIN VALLEY, CALIFORNIA
Sponsoring Institution
National Institute of Food and Agriculture
Project Status
ACTIVE
Funding Source
Reporting Frequency
Annual
Accession No.
1031069
Grant No.
2023-68016-40683
Cumulative Award Amt.
$299,440.00
Proposal No.
2023-04667
Multistate No.
(N/A)
Project Start Date
Aug 15, 2023
Project End Date
Aug 14, 2025
Grant Year
2023
Program Code
[A1712]- Rapid Response to Extreme Weather Events Across Food and Agricultural Systems
Recipient Organization
UNIV OF CALIFORNIA
(N/A)
SANTA BARBARA,CA 93106
Performing Department
(N/A)
Non Technical Summary
In early 2023, California received extreme levels of precipitation, with a series of atmospheric rivers. Numerous rivers including the Kern peaked at heights not seen in 50y. Furthermore, the Sierra Nevada mountains, which drain to the San Joaquin Valley (SJV), have received more than 50' of snowfall, breaking a 40y record. A state of emergency was declared in 47 of 58 counties with a Presidential Major Disaster Declaration issued on 4/3/2023 for several key agricultural producing counties including Kern County, the focus of this integrated project.The enormous increase in water availability including widespread standing water and ongoing floods will dramatically change the ecology of the SJV's agroecosystems, with potentially important implications for farmworker health, well-being and safety and agroecosystem resilience.We propose a multidisciplinary project, leveraging remote sensing, ecology and human geography approaches, to evaluate (1) how West Nile virus risk and mosquito abundance is responding spatially and temporally to increased standing water, and how farmworkers can adapt to changes in risk (2) how farmers are responding to the flood via planting and fallowing decisions, and whether their responses are likely to improve or degrade agroecosystem resilience in future extreme weather events.
Animal Health Component
20%
Research Effort Categories
Basic
70%
Applied
20%
Developmental
10%
Classification

Knowledge Area (KA)Subject of Investigation (SOI)Field of Science (FOS)Percent
8070420107040%
7220420107030%
8070420308030%
Goals / Objectives
In early 2023, California received extreme levels of precipitation, with a series of atmospheric rivers. Numerous rivers including the Kern peaked at heights not seen in 50y. Furthermore, the Sierra Nevada mountains, which drain to the San Joaquin Valley (SJV), have received more than 50' of snowfall, breaking a 40y record. A state of emergency was declared in 47 of 58 counties with a Presidential Major Disaster Declaration issued on 4/3/2023 for several key agricultural producing counties including Kern County, the focus of this integrated project.The enormous increase in water availability including widespread standing water and ongoing floods will dramatically change the ecology of the SJV's agroecosystems, with potentially important implications for farmworker health, well-being and safety and agroecosystem resilience.In this multidisciplinary project, we will leverage remote sensing, ecology and human geography approaches, to evaluate (1) how West Nile virus risk and mosquito abundance is responding spatially and temporally to increased standing water, and how farmworkers can adapt to changes in risk (2) how farmers are responding to the flood via planting and fallowing decisions, and whether their responses are likely to improve or degrade agroecosystem resilience in future extreme weather events.The specific objectives of this project are to:1: Quantify spatiotemporal patterns of abiotic conditions and cropping patterns.2: Build predictive models of WNV suitability and response to flooding.3: Engage producers in understanding the impacts of climate extremes, such as flooding, to agriculture in the region and co-develop management recommendations to improve climate resilience.4: Conduct extension activities to promote education and understanding of potential risks and adaptive responses to flooding and extreme weather variability.
Project Methods
O1: Quantify spatiotemporal patterns of abiotic conditions and cropping patterns.Agricultural land use including cropping patterns and the amount and distribution of idle lands will be ascertained from field-level vector data from the Kern County Agricultural Commissioner's office. These annual shapefiles include information on the farmer, crop type, dates active and inactive, which allows us to track agricultural land use over time. Annual data up to 2022 are available, which allows us to determine what farms, fields and crops, per 2022, are affected by flooding based on the remote sensing data.We have already developed methodology for identifying retired lands using the 1997-2021 annual data, which we can extend to 2022-23.We will supplement these geospatial crop records with satellite imagery so we can identify changes in land retirement in 2023 to feed into the WNV model. Specifically, we will map both short-term fallowing and long-term retirement using seasonal patterns in parcel-level land cover. We will characterize land cover using cross-calibrated spectral mixture models developed by Sousa, and to Landsatand Sentinel-2 data. A key benefit of using this approach is that it will allow us to quantify near-real time status of 2023 surface water (flooding) distribution, soil moisture and vegetation cover, in addition to creating a retired/cropped binary classification. We will further link these unmixing-derived maps to EVI2 and NDWI spectral indices from near-daily 3.15m resolution PlanetScope data using scaling relationshipsto develop a fused product at high spatial and temporal resolution for the 2023 growing season. These resulting biophysical maps will then complement our producer interviews in further understanding how surface water availability affected planting decisions county-wide in 2023. Moreover,near-real time mapping of surface water will be critical to understanding how mosquito sources and abundance are likely to change throughout the year, and will be a critical input into WNV models in O2.O2: Build predictive model of WNV suitability and response to flooding.To model current WNV activity in mosquito populations, at high spatial and temporal resolution in Kern County, we will use boosted regression trees (BRT),with positive WNV mosquito "pools" (groups of mosquitos tested together) from the Mosquito and Vector Control Association of California (MVCAC) and Kern Mosquito and Vector Control District's weekly surveillance data as the outcome variable. Boosted regression trees are a machine-learning based approach to modeling environmental suitability for target species. Importantly, BRT allows for abundance estimates (e.g., number of WNV positive mosquito pools, or minimum infection rate) and true zero observations to be incorporated into models, rather than occurrence alone.We will model WNV positive mosquito pools as a function of mosquito abundance, passerine bird host abundance from ebird status and trends data, as well as abiotic conditions (e.g., temperature, ET, surface water) from remote sensing in O1, and additional land cover data (as in O1 above). We will model WNV risk based on past surface water conditions, and forecast to current flooding conditions to make predictions of how WNV risk will change spatially and temporally in response to the extreme rain year. We will then assess model performance based on the coming year's WNV surveillance data, and update our BRT training and testing datasets to incorporate this new surveillance data to improve model performance in predicting future responses to extreme weather events. Finally, we will assess the spatial and temporal overlap between high WNV risk and at-risk human populations and communities within the greater Bakersfield region and Kern County more broadly.O3: Engage producers in understanding the impacts of climate extremes, such as flooding, to agriculture in the region and co-develop management recommendations to improve climate resilience.To understand producer perspectives, we will conduct qualitative semi-structured interviews. Interviewees will be selected via purposive, snowball sampling design methods, which are commonly used in intensive case studies that involve hard-to-find populations. Thus, using snowball sampling techniques, we will contact stakeholders already within our professional networks, and then ask interviewees to recommend producers and land owners within their networks that we could contact. We aim for 10-15 interviews, a sample size that is appropriate and effective to uncover the major themes and topics relevant to the research topic and with all key stakeholder groups. The interviews aim to understand the perspectives of stakeholders in depth, instead of covering the entire breadth of perspectives and experiences that likely exist.O4: Conduct extension activities to promote education and understanding of potential risks and adaptive responses to flooding and weather swings ("Efforts" that will be undertaken to cause a change in knowledge, actions and conditions). The goal of the extension activities are three-fold: 1) rapidly provide actionable information to the Kern Mosquito and Vector Control District, University of California Cooperative Extension, and California Department of Public Health to improve surveillance and management of changing West Nile virus hotspots of suitability, 2) increase citizen engagement and public knowledge through participation of local producers and stakeholders in the research and extension process, and 3) build science-based capability and technical competency in undergraduate and professional master's students at Hispanic Serving Institutions (UCSB, SDSU) through informal and formal curriculum development and internship experience in ecology, remote sensing and qualitative methods.Evaluation of project success: Methods for evaluating success of research activities are described above (e.g., assessment of WNV model performance using updated surveillance data); feedback from the academic community through conferences and peer review will also be used to evaluate success of research activities. Evaluation of extension success will be based on development of relationships with stakeholders (Kern MVCD, CDPH, UCANR extension), co-development of extension objectives with these stakeholders, and implementation of co-developed outreach materials (e.g., web-based interactive WNV risk mapping) with the goal of informing decision making in agricultural management decisions and vector control activities and public health interventions in the near-term.

Progress 08/15/23 to 08/14/24

Outputs
Target Audience:The target audience reached by our efforts during this reporting period include students, public health and vector control district employees and personnel, stakeholders including farmers and water managers, researchers, and members of the general public in Kern County, CA. Students include undergraduate and graduate students at SDSU who participated in research as well as took courses in which material generated by this grant was taught or used for instruction, professional masters and PhD students in the Bren School at UCSB who both participated in research as well as took courses in which material generated by this grant was taught or used for instruction. Public health and vector control district personnel include employees of the Kern Mosquito and Vector Control District, with whom we have developed a close collaboration, leadership in the Vector-Borne Disease Section of the California Department of Public Health (CDPH), as well as state and local epidemiologists from CDPH, Santa Clara County Department of Health, LA County Department of Health, Pasadena City Department of Health and Long Beach Department of Health. Other stakeholders include farmers and water managers in Kern County in the Central Valley, including some who participated in interviews for this project. Changes/Problems:The main challenge that we have faced in this reporting period is that our contact at Kern UCANR, who wrote a letter of support for this work, retired early on in the grant period. They have not yet filled this position, so our original outreach and extension plans in collaboration with UCANR have been set back. We have since made new contacts at UCANR in Kern and plan to leverage those to achieve our original goals for outreach and extension. What opportunities for training and professional development has the project provided?In the reporting period, we trained multiple undergraduate and graduate students, as well as post-bac scholars in intensive, multi-disciplinary research and engagement with stakeholders. We trained one PhD student and post-bac scholar in disease ecology, geospatial methods and scientific writing. These scholars also participated in several of the calls with the vector control district personnel. We also trained one post-master's student in app development via Rshiny, at no cost to the grant. We trained five SDSU undergraduates, one master's and one PhD student in qualitative methods-conducting and/or analyzing qualitative interviews and newspaper content analysis.At SDSU, we also trained 5 undergraduate students and 3 graduate students in multispectral and hyperspectral remote sensing methodology. This included the use of open-source Python packages for image geolocation and orthorectification; evaluating time series of flooding and interpreting spatiotemporal patterns; and producing map products in ArcGIS Pro. How have the results been disseminated to communities of interest?Results are being disseminated to academic researchers, practitioners and the public via communications targeting all three groups. We presented results from this project at 3 conferences including the Ecological Society of America in 2023 (invited talk) and 2024 (contributed talk) and Ecology and Evolution of Infectious Disease in 2024 (invited talk and contributed poster). Four manuscripts are currently submitted or under peer review at Water Policy, Agricultural Water Management, Ecology Letters and PNAS. We have had multiple calls with Vector Control Districts to share results, communicate data gaps and develop collaborations to apply our expertise to local challenges facing rural communities. We created an Rshiny visual interface that was shared with the Kern Vector Control District to facilitate the communication of mosquito data and mosquito-borne disease risk with vector control personnel and the public. We have also communicated results to the California Department of Public Health and developed further collaborations with state and local public health departments. We also produced a commentary to reach a public audience. What do you plan to do during the next reporting period to accomplish the goals?In the remaining No Cost Extension period, we intend to: Continue evaluating how the 2023 flooding impacted farmers with respect to crop composition, timing of planting, and pesticide use. We anticipate a manuscript will be submitted by January 2025 led by co-PI Larsen. We will continue developing collaborations with the vector control districts and local public health departments to further evaluate mosquito-borne disease risk response to environmental conditions. Communicate results and co-develop outreach with the Kern County UCANR office. Archive data from the project on a public repository such as dryad or github, along with code needed to reproduce the findings of our publications.

Impacts
What was accomplished under these goals? In the reporting period, we achieved the majority of our goals. While research products are still in the peer-review pipeline, we successfully quantified the impact of abiotic conditions, including through standing water remote sensing (MacDonald et al., submitted) and long-term drought fluctuations (Sambado et al. submitted) on West Nile Virus (goals 1,2). We also developed and validated a West Nile virus risk mapping approach (MacDonald et al. 2024). We conducted and analyzed interviews and newspaper articles (Quandt et al. in review), wrote public-facing commentaries, developed an Rshiny web interface for communicating results to stakeholders and the general public, engaged with the vector control district in the study region and developed curricula for several quantitative and qualitative courses at the undergraduate and master's level (goals 3,4).

Publications

  • Type: Peer Reviewed Journal Articles Status: Published Year Published: 2024 Citation: MacDonald, AJ, D Hyon, S Sambado, K Ring & A Boser, 2024. Remote sensing of temperature-dependent mosquito and viral traits predicts field surveillance-based disease risk. Ecology, 105(11):e4420.
  • Type: Peer Reviewed Journal Articles Status: Published Year Published: 2024 Citation: Mattson, M, D Sousa, A Quandt, P Ganster and T Biggs, 2024. Mapping multi-decadal wetland loss: Comparative analysis of linear and nonlinear spatiotemporal characterization. Remote Sensing of Environment, 302:113969.
  • Type: Peer Reviewed Journal Articles Status: Published Year Published: 2024 Citation: Small, C and D Sousa, 2024. The Standardized Spectroscopic Mixture Model. Remote Sensing, 16(20):3768.
  • Type: Peer Reviewed Journal Articles Status: Published Year Published: 2024 Citation: Small, C and D Sousa, 2024. Robust Cloud Suppression and Anomaly Detection in Time-Lapse Thermography. Remote Sensing, 16(2):255.
  • Type: Peer Reviewed Journal Articles Status: Published Year Published: 2024 Citation: Small, C and D Sousa, 2024. Spectroscopic Phenological Characterization of Mangrove Communities. Remote Sensing, 16(15):2796.
  • Type: Peer Reviewed Journal Articles Status: Published Year Published: 2024 Citation: Rogers, JA, KM Robertson, TJ Hawbaker and D Sousa, 2024. Classifying Plant Communities in the North American Coastal Plain With PRISMA Spaceborne Hyperspectral Imagery and the Spectral Mixture Residual. Journal of Geophysical Research: Biogeosciences, 129(9):e2024JG008217.
  • Type: Peer Reviewed Journal Articles Status: Published Year Published: 2024 Citation: Miner, K, L Baskaran, B Gay, D Sousa and C Miller, 2024. Frozen no more, a case study of Arctic permafrost impacts of oil and gas withdrawal. Sci Rep 14, 25403.